{
  "service": "mineru",
  "displayName": "MinerU",
  "categories": [
    "AI",
    "Productivity"
  ],
  "authTypes": [
    "api_key"
  ],
  "auth": [
    {
      "type": "api_key",
      "label": "API Token",
      "placeholder": "mineru_api_token",
      "description": "MinerU API token used with the Authorization Bearer header. Apply for or manage it here: https://mineru.net/apiManage/token"
    }
  ],
  "homepageUrl": "https://mineru.net",
  "actions": [
    {
      "id": "mineru.create_extract_batch",
      "service": "mineru",
      "name": "create_extract_batch",
      "description": "Create a MinerU precise extraction batch from document URLs.",
      "requiredScopes": [],
      "providerPermissions": [],
      "inputSchema": {
        "type": "object",
        "properties": {
          "files": {
            "type": "array",
            "items": {
              "type": "object",
              "properties": {
                "url": {
                  "type": "string",
                  "description": "The public HTTP or HTTPS URL of one document to parse with MinerU.",
                  "format": "uri"
                },
                "is_ocr": {
                  "type": "boolean",
                  "description": "Whether to enable OCR for this file. Only applies to pipeline and vlm models."
                },
                "data_id": {
                  "type": "string",
                  "minLength": 1,
                  "maxLength": 128,
                  "description": "The caller-defined data identifier associated with the source document."
                },
                "page_ranges": {
                  "type": "string",
                  "minLength": 1,
                  "description": "The page range expression to parse for this file."
                }
              },
              "additionalProperties": false,
              "required": [
                "url"
              ],
              "description": "One URL-based MinerU batch extraction file descriptor."
            },
            "description": "The URL-based documents to submit for batch parsing.",
            "minItems": 1,
            "maxItems": 50
          },
          "enable_formula": {
            "type": "boolean",
            "description": "Whether to enable formula recognition for pipeline and vlm models."
          },
          "enable_table": {
            "type": "boolean",
            "description": "Whether to enable table recognition for pipeline and vlm models."
          },
          "language": {
            "type": "string",
            "minLength": 1,
            "description": "The document language code used by pipeline and vlm models."
          },
          "extra_formats": {
            "type": "array",
            "items": {
              "type": "string",
              "enum": [
                "docx",
                "html",
                "latex"
              ],
              "description": "An additional export format requested for the parsed result."
            },
            "description": "Additional result formats to export besides Markdown and JSON.",
            "minItems": 1
          },
          "model_version": {
            "type": "string",
            "enum": [
              "pipeline",
              "vlm",
              "MinerU-HTML"
            ],
            "description": "The MinerU model version to use for parsing."
          },
          "no_cache": {
            "type": "boolean",
            "description": "Whether MinerU should bypass its URL content cache."
          },
          "cache_tolerance": {
            "type": "integer",
            "minimum": 0,
            "description": "The acceptable MinerU URL cache age in seconds when no_cache is false."
          }
        },
        "additionalProperties": false,
        "required": [
          "files"
        ],
        "description": "Input payload for creating a MinerU URL batch extraction task."
      },
      "outputSchema": {
        "type": "object",
        "properties": {
          "batch_id": {
            "type": "string",
            "description": "The MinerU batch extraction task identifier."
          },
          "trace_id": {
            "type": "string",
            "description": "The upstream request trace identifier."
          },
          "msg": {
            "type": "string",
            "description": "The upstream response message."
          }
        },
        "additionalProperties": false,
        "required": [
          "batch_id"
        ],
        "description": "The normalized output payload for creating a MinerU extraction batch."
      }
    },
    {
      "id": "mineru.create_extract_task",
      "service": "mineru",
      "name": "create_extract_task",
      "description": "Create a MinerU precise extraction task from a document URL.",
      "requiredScopes": [],
      "providerPermissions": [],
      "inputSchema": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "description": "The public HTTP or HTTPS URL of the document to parse with MinerU.",
            "format": "uri"
          },
          "is_ocr": {
            "type": "boolean",
            "description": "Whether to enable OCR. Only applies to pipeline and vlm models."
          },
          "enable_formula": {
            "type": "boolean",
            "description": "Whether to enable formula recognition for pipeline and vlm models."
          },
          "enable_table": {
            "type": "boolean",
            "description": "Whether to enable table recognition for pipeline and vlm models."
          },
          "language": {
            "type": "string",
            "minLength": 1,
            "description": "The document language code used by pipeline and vlm models."
          },
          "data_id": {
            "type": "string",
            "minLength": 1,
            "maxLength": 128,
            "description": "The caller-defined data identifier associated with the source document."
          },
          "extra_formats": {
            "type": "array",
            "items": {
              "type": "string",
              "enum": [
                "docx",
                "html",
                "latex"
              ],
              "description": "An additional export format requested for the parsed result."
            },
            "description": "Additional result formats to export besides Markdown and JSON.",
            "minItems": 1
          },
          "page_ranges": {
            "type": "string",
            "minLength": 1,
            "description": "The page range expression to parse, such as 2,4-6 or 2--2."
          },
          "model_version": {
            "type": "string",
            "enum": [
              "pipeline",
              "vlm",
              "MinerU-HTML"
            ],
            "description": "The MinerU model version to use for parsing."
          },
          "no_cache": {
            "type": "boolean",
            "description": "Whether MinerU should bypass its URL content cache."
          },
          "cache_tolerance": {
            "type": "integer",
            "minimum": 0,
            "description": "The acceptable MinerU URL cache age in seconds when no_cache is false."
          }
        },
        "additionalProperties": false,
        "required": [
          "url"
        ],
        "description": "Input payload for creating a MinerU extraction task from a document URL."
      },
      "outputSchema": {
        "type": "object",
        "properties": {
          "task_id": {
            "type": "string",
            "description": "The MinerU extraction task identifier."
          },
          "trace_id": {
            "type": "string",
            "description": "The upstream request trace identifier."
          },
          "msg": {
            "type": "string",
            "description": "The upstream response message."
          }
        },
        "additionalProperties": false,
        "required": [
          "task_id"
        ],
        "description": "The normalized output payload for creating a MinerU extraction task."
      }
    },
    {
      "id": "mineru.get_extract_batch_results",
      "service": "mineru",
      "name": "get_extract_batch_results",
      "description": "Get the current status and result URLs for a MinerU extraction batch.",
      "requiredScopes": [],
      "providerPermissions": [],
      "inputSchema": {
        "type": "object",
        "properties": {
          "batch_id": {
            "type": "string",
            "minLength": 1,
            "description": "The MinerU batch extraction task identifier."
          }
        },
        "additionalProperties": false,
        "required": [
          "batch_id"
        ],
        "description": "Input payload for looking up MinerU batch extraction results."
      },
      "outputSchema": {
        "type": "object",
        "properties": {
          "batch_id": {
            "type": "string",
            "description": "The MinerU batch extraction task identifier."
          },
          "extract_result": {
            "type": "array",
            "items": {
              "type": "object",
              "properties": {
                "file_name": {
                  "type": "string",
                  "description": "The submitted source file name, when returned."
                },
                "data_id": {
                  "type": "string",
                  "description": "The caller-defined data identifier, when provided."
                },
                "state": {
                  "type": "string",
                  "description": "The upstream task state, such as waiting-file, pending, running, converting, done, or failed."
                },
                "full_zip_url": {
                  "type": "string",
                  "description": "The ZIP URL containing Markdown, JSON, and any requested exports."
                },
                "err_msg": {
                  "type": "string",
                  "description": "The upstream failure reason when the task failed."
                },
                "extract_progress": {
                  "type": "object",
                  "properties": {
                    "extracted_pages": {
                      "type": "integer",
                      "description": "The number of pages that have already been parsed."
                    },
                    "total_pages": {
                      "type": "integer",
                      "description": "The total number of pages reported for the document."
                    },
                    "start_time": {
                      "type": "string",
                      "description": "The upstream parsing start time."
                    }
                  },
                  "additionalProperties": false,
                  "description": "The upstream parsing progress details returned while the task is running."
                }
              },
              "additionalProperties": false,
              "required": [
                "state"
              ],
              "description": "The normalized MinerU batch extraction item result."
            },
            "description": "The extraction results for files in the requested batch."
          },
          "trace_id": {
            "type": "string",
            "description": "The upstream request trace identifier."
          },
          "msg": {
            "type": "string",
            "description": "The upstream response message."
          }
        },
        "additionalProperties": false,
        "required": [
          "batch_id",
          "extract_result"
        ],
        "description": "The normalized output payload for MinerU batch extraction results."
      }
    },
    {
      "id": "mineru.get_extract_task",
      "service": "mineru",
      "name": "get_extract_task",
      "description": "Get the current status and result URLs for a MinerU extraction task.",
      "requiredScopes": [],
      "providerPermissions": [],
      "inputSchema": {
        "type": "object",
        "properties": {
          "task_id": {
            "type": "string",
            "minLength": 1,
            "description": "The MinerU extraction task identifier."
          }
        },
        "additionalProperties": false,
        "required": [
          "task_id"
        ],
        "description": "Input payload for looking up a MinerU extraction task."
      },
      "outputSchema": {
        "type": "object",
        "properties": {
          "task_id": {
            "type": "string",
            "description": "The MinerU extraction task identifier."
          },
          "data_id": {
            "type": "string",
            "description": "The caller-defined data identifier, when provided."
          },
          "state": {
            "type": "string",
            "description": "The upstream task state, such as pending, running, converting, done, or failed."
          },
          "full_zip_url": {
            "type": "string",
            "description": "The ZIP URL containing Markdown, JSON, and any requested exports."
          },
          "err_msg": {
            "type": "string",
            "description": "The upstream failure reason when the task failed."
          },
          "extract_progress": {
            "type": "object",
            "properties": {
              "extracted_pages": {
                "type": "integer",
                "description": "The number of pages that have already been parsed."
              },
              "total_pages": {
                "type": "integer",
                "description": "The total number of pages reported for the document."
              },
              "start_time": {
                "type": "string",
                "description": "The upstream parsing start time."
              }
            },
            "additionalProperties": false,
            "description": "The upstream parsing progress details returned while the task is running."
          }
        },
        "additionalProperties": false,
        "required": [
          "task_id",
          "state"
        ],
        "description": "The normalized MinerU extraction task result."
      }
    }
  ]
}